WP 26-011: Data Centers and Rural Environmental Disamenities: Issues in Indian Country

Data centers, built rapidly to process artificial-intelligence workloads, impose local costs while their benefits accrue largely elsewhere: higher electricity prices, heavy water use for cooling, and, through fossil-fuel backup power and grid demand, degraded air quality and added emissions. Some American Indian tribes argue that these facilities disproportionately burden Indian Country, drawing on shared water and infrastructure without returning jobs or economic gains. Using an environmental-justice framework and two spatial approaches, a land-based counterfactual and a leading cost-optimization siting model, we ask where U.S. data centers sit relative to tribal lands and where such a model would place future ones. Data centers do not broadly target tribal land at this early stage. Conditioned on region, both existing centers and the model’s most suitable sites are, if anything, less Native than the surrounding land. But the burden is more specific and likely to build over time: where data centers and tribal communities coincide, they disproportionately share water-stressed watersheds, especially in the Southwest. The cost-minimizing criteria steering siting (rurality, low land value, low taxes, distance from developed land, and political weakness) are the same criteria that routed dams, mines, and waste to tribal lands, suggesting future siting may burden them disproportionately.